UPDATED 09:00 EST / MAY 19 2020

INFRA

Xilinx’s new smaller programmable chips enable machine learning in space

Xilinx Inc. today launched what it says is the industry’s first-ever 20-nanometer, space-grade field-programmable gate array, which is a special kind of computer chip built on an integrated circuit that can be reprogrammed for specific computing tasks.

The new 20nm Radiation Tolerant Kintex UltraScale XQRKU060 FPGA is designed for use in orbital satellites, and delivers a significant reduction in size, weight and power compared with the company’s previous, 65nm space-grade chips. It also comes equipped with “robust, radiation-tolerant features,” the company said.

The chip provides a 10-times increase in digital signal processing performance, while also enabling machine learning to be performed in outer space for the first time, Xilinx said.

“The 20nm RT Kintex UltraScale FPGA is breaking industry standards and setting a new benchmark for meeting the high compute requirements of high bandwidth payloads, space exploration and research missions,” said Minal Sawant, a system architect and space products manager at Xilinx.

Xilinx’s chips have long played a crucial role in space exploration. Its hardware has previously been used to power missions such as NASA’s Mars Rovers and the Solar Orbiter that’s currently headed toward the sun.

The new XQRKU060 chip enables some important new capabilities for satellites, including the ability to update in real-time, deliver video-on-demand and perform compute processes on the fly, the company said. It also supports the TensorFlow and PyTorch machine learning frameworks, enabling satellites to perform advanced tasks such as scientific analysis, object detection and image classification.

The XQRKU060 FPGA can be programmed in orbit with Xilinx’s newly updated Vivado Design Suite, which helps to eliminate routing congestion. That ensures that 90% of the chip’s processing capabilities can be dedicated to the tasks it’s assigned to do without any performance degradation.

“Machine learning is becoming the pervasive approach to solving problems for which conventional programming approaches fail,” Moor Insights & Strategy senior analyst Karl Freund told SiliconANGLE. “Consequently,  a space-based platform is a natural next step, and it would seem that Xilinx is ready to begin that journey.”

Xilinx said mechanical samples and prototype units of the XQRKU060 FPGA are available now, while flight units will be available in September. Customers can start prototyping with the KCU105 Evaluation Kit or the Kintex UltraScale Space Development Kit immediately.

Image: PIRO4D/Pixabay

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU